New research on relative risks of different modes.

User avatar
[XAP]Bob
Posts: 18065
Joined: 26 Sep 2008, 4:12pm

Re: New research on relative risks of different modes.

Postby [XAP]Bob » 8 Dec 2012, 12:22pm

Given that we're looking for the relative risks of travelling along roads, comparing cyclists, pedestrians and motorists I think it's perfectly appropriate to exclude motorways (where two classes of user are excluded) and cyclepaths (where one class is excluded) as well as off road courses (where at least one class is excluded)

The interesting thing to compare is like-for-like data - i.e. collected over the same period at the same locations. Saying that inter-junction motorway driving is safer than throwing yourself off a mountain is hardly surprising.

Saying that pedestrians (not fell runners) are more likely to be killed by a car than a cyclist covering the same journey is VERY significant.
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.

User avatar
CJ
Posts: 3119
Joined: 15 Jan 2007, 9:55pm

Re: New research on relative risks of different modes.

Postby CJ » 10 Dec 2012, 4:01pm

Ellieb wrote:
Correcting for demographics is also important. Young males are disproportionately likely to get involved in crashes, whatever mode of transport they happen to be using, yet form a disproportionate proportion of cyclists.


Really? Where do you get that stat?

I don't know where THEY got that stat, but the National Travel Survey found that the average teenager makes the greatest number of walking and cycling trips per annum and that this number falls gradually but steadily with age, such that those in their 50s make about 1/3 fewer active trips compared to the under 17s. 'Disporportionate' is nevertheless, perhaps too strong a word for a ratio of 3 to 2.
Chris Juden
One lady owner, never raced or jumped.

Ellieb
Posts: 899
Joined: 26 Jul 2008, 7:06pm

Re: New research on relative risks of different modes.

Postby Ellieb » 10 Dec 2012, 5:08pm

As you say. This includes walking as well as cycling and ignores the stat that shows older people cycling a greater distance.

Pete Owens
Posts: 1938
Joined: 7 Jul 2008, 12:52am

Re: New research on relative risks of different modes.

Postby Pete Owens » 11 Dec 2012, 12:56am

If you look at tables 3 & 4 of the report you see that young males (<30) account for about 10% of total car use but 22% of cycle use. So the most risky demographic is over-represented in the cycling population by a factor of 2.

Yes, 30-50s cycle a bit more than under 30s, but they drive very much further.

User avatar
CJ
Posts: 3119
Joined: 15 Jan 2007, 9:55pm

Re: New research on relative risks of different modes.

Postby CJ » 11 Dec 2012, 10:15am

Ellieb wrote:But even on a casual reading things like trying to take out motorway mileages: One of the points they make is that people generally commute for the same amount of time regardless of mode & that is why they prefer to measure the risk in terms of per hour. Well, an awful lot of people commute on motorways and do so for periods of less than an hour. To try to claim that motorway mileage is solely a sort of 'long duration' travel & should not be compared to cycling I would suggest is a distortion.

Where in the report do the authors subtract motorway mileage? I cannot find any evidence that they did that. All I can find is an observation that comparing risks per hour rather than per mile reduces the distortion caused by very long distance (motorway) trips, for which cycling is not an option.

Ellieb wrote:Likewise they seek to eliminate 'high risk' activities like BMX/sportives etc from the cycling stats on the grounds it isn't transport. I would suggest that a significant source of young male car fatalities isn't 'transport' either. Racing your mates or impressing the girls by doing tricks & stunts accounts for an awful lot of late night accidents. It is using a vehicle to have fun...just like mountain biking.

They seek to eliminate mountain-biking and BMX alright, by excluding off-highway cycling fatalities, but cannot avoid including some off-highway falls due to V19.8 being a 'dustbin code', given to cycling falls in unknown locations.

Nowhere does it say that they even tried to exclude sportives and the like. And there's no way they could have done that even if they'd wanted to, since there's not a separate ICD code for sport cycling crashes - they all get coded as transport. We know that on-road sport cycling is more hazardous than riding for travel and transport. And given the tendency of sporty types to use their cycling commute for training purposes, even transport cycling fatalities will be inflated by the greater risks that may be taken by cyclists who are unusually motivated to maintain their speed and rythum. There's plenty of evidnece on U-tube of how risky that is; and I'm not sorry to say that if I were to wear a camera, my commute wouldn't make such exciting viewing - more Countryfile than Road-Wars!

When I consider the preponderance of sporty cyclists amongst the tiny number of cycle users in UK, compared to the Netherlands where this type is swamped by an overwhelming majority of normally risk-averse people on bikes, I cannot help but wonder how far the danger-sport factor goes towards making cycling four times more lethal here than there. Meanwhile I wonder less at my good fortune - never to have been hurt whilst pootling to work on my mostly quiet back road route.
Chris Juden
One lady owner, never raced or jumped.

User avatar
meic
Posts: 19355
Joined: 1 Feb 2007, 9:37pm
Location: Caerfyrddin (Carmarthen)

Re: New research on relative risks of different modes.

Postby meic » 11 Dec 2012, 10:23am

And given the tendency of sporty types to use their cycling commute for training purposes, even transport cycling fatalities will be inflated by the greater risks that may be taken by cyclists who are unusually motivated to maintain their speed and rythum.


In my experience a fair few of the motoring crashes are due to similar minded car drivers and motorcyclists, I am not sure this is more significant among cyclists. Those involved will point out that they are much more skilled than other road users.
Yma o Hyd

User avatar
CJ
Posts: 3119
Joined: 15 Jan 2007, 9:55pm

Re: New research on relative risks of different modes.

Postby CJ » 11 Dec 2012, 12:04pm

meic wrote:
And given the tendency of sporty types to use their cycling commute for training purposes, even transport cycling fatalities will be inflated by the greater risks that may be taken by cyclists who are unusually motivated to maintain their speed and rythum.


In my experience a fair few of the motoring crashes are due to similar minded car drivers and motorcyclists, I am not sure this is more significant among cyclists. Those involved will point out that they are much more skilled than other road users.

Very few drivers - apart from those driving emergency vehicles - ever drive as fast as they possibly could, whereas a sports-minded cyclist will be doing that rather often, sometimes in a state of near exhaustion. And most people think they're more skilled than the average road user.

Fact is: 13% of the cyclists in Dutch hospitals are helmetted roadies and mountain-bikers, in spite of comprising only about 1% of Dutch cycle traffic. Maybe they are more skilfull, but if that's the case they're apparently consuming any potential benefit and then some, in the hugely greater risks they must be taking. It's the only way that I can think of reconciling those numbers.
Chris Juden
One lady owner, never raced or jumped.

User avatar
[XAP]Bob
Posts: 18065
Joined: 26 Sep 2008, 4:12pm

Re: New research on relative risks of different modes.

Postby [XAP]Bob » 11 Dec 2012, 12:28pm

Virtually everyone claims to be a better than average driver, but that's no surprise.

I also try to maintain rhythm in the car, I'll slow in advance of joining a queue (just lift foot) so that I have to do less braking, and often leas accelerating (the queue often evaporates as I arrive).

That is different from the full speed rhythm that induces people through blind gaps and junctions.
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.

User avatar
meic
Posts: 19355
Joined: 1 Feb 2007, 9:37pm
Location: Caerfyrddin (Carmarthen)

Re: New research on relative risks of different modes.

Postby meic » 11 Dec 2012, 12:31pm

Yes to some extent the helmets in Holland help you to identify a type of rider.
Somebody on a motorcycle, especially a large powerful one*, is more likely to be after a bit of road sport and this shows up in their accident stats.
However with cars the boy-racers are generally hidden within the whole of car users and this makes them invisible as a sub-group but my own experience says that they are still there and with a racer's attitude and with a racer's accident rate.
The insurance companies agree with me as my insurance is ten to twenty times less than theirs!

*I know that there are "safe" riders of superbikes out there, I am one :wink: but when I first started I would have given the Alley-cats a run for their money. :oops:
Yma o Hyd

User avatar
RickH
Posts: 5324
Joined: 5 Mar 2012, 6:39pm
Location: Horwich, Lancs.

Re: New research on relative risks of different modes.

Postby RickH » 11 Dec 2012, 12:38pm

CJ wrote:Fact is: 13% of the cyclists in Dutch hospitals are helmetted roadies and mountain-bikers, in spite of comprising only about 1% of Dutch cycle traffic. Maybe they are more skilfull, but if that's the case they're apparently consuming any potential benefit and then some, in the hugely greater risks they must be taking. It's the only way that I can think of reconciling those numbers.

It depends on several factors, particularly if the statistics include all cyclists hospitalised - including ones injured during events where helmet wearing is compulsory (road race mass pile-ups for example as we see each year on the TdF) rather than limited to "traffic" incidents - then the statistic will be skewed possibly to a very large extent by "event" injuries, not even happening on the road in the case of mountainbiking/cyclocross events, rather than "training" ones.

Rick.

User avatar
CJ
Posts: 3119
Joined: 15 Jan 2007, 9:55pm

Re: New research on relative risks of different modes.

Postby CJ » 12 Dec 2012, 2:29pm

RickH wrote:
CJ wrote:Fact is: 13% of the cyclists in Dutch hospitals are helmetted roadies and mountain-bikers, in spite of comprising only about 1% of Dutch cycle traffic. Maybe they are more skilfull, but if that's the case they're apparently consuming any potential benefit and then some, in the hugely greater risks they must be taking. It's the only way that I can think of reconciling those numbers.

It depends on several factors, particularly if the statistics include all cyclists hospitalised - including ones injured during events where helmet wearing is compulsory (road race mass pile-ups for example as we see each year on the TdF) rather than limited to "traffic" incidents - then the statistic will be skewed possibly to a very large extent by "event" injuries, not even happening on the road in the case of mountainbiking/cyclocross events, rather than "training" ones.

I'm sure you're right, that 13% most probably includes injuries sustained whilst actually racing, including off-road and even in velodromes. As for whether those injuries account "to a very large extent" for the apparent hazards of sporty cycling, I doubt it. The discrepancy between 1% and 13% is too big for all of it to be heaped onto the tiny fraction of one percent that actual competition contributes to the sum total of Dutch cycling. That discrepancy nevertheless deserves further study and I hope someone in the Nethelands may eventually report in more detail. Until they do so, I think that it's a sensible working hypothesis that sporty cycling is a few orders of magnitude more hazardous than normal cycling and that the burden of proof lies with those who might pretend otherwise.

I look at it this way. The Dutch cycling population is like the British driving population: in both cases 'risky racers' are a small minority, diluted by the vast majority of normally risk-averse people. But the British cycling population is a 'boiled down' minority, in which the sporty fraction have achieved dominance by evaporation of the majority to modes of transport deemed more socially acceptable in English-speaking cultures.

And it's easy to see why more athletic cycling may be more hazardous. Speed alone increases the problem with every driver who assumes cycles to be slow (or easy to bully) and pulls out or across us. And above a certain level, more effort reduces concentration. (CP Davey - Ergonomics 1973 - found that whilst moderate exertion improves mental processes, higher and prolonged effort leads to mistakes. The expression: 'eyeballs out' perhaps says it all!) Meanwhile the need for speed and distance leads sporty cyclists onto smoother and straighter roads, which generally carry more and faster traffic.

It wouldn't do to overstate this effect, but it follows that the apparently greater risk of cycling in Britain is likely to be due in part, to the elimination of naturally safer cyclists from the sport-dominated UK cycling scene. So cycling for any less energetic types who can nevertheless be persuaded to do it here, is likely to be significantly better than is suggested by the overall figure.
Chris Juden
One lady owner, never raced or jumped.

Brucey
Posts: 43043
Joined: 4 Jan 2012, 6:25pm

Re: New research on relative risks of different modes.

Postby Brucey » 12 Dec 2012, 3:21pm

this may sound like a daft question, but how is (say) '1% of cycle traffic' calculated?

Per rider?
Per mile ridden?
Per Hour ridden?
Per Journey?

It has been my observation that the 'racey types' also put the miles in for training, often under simulated race conditions, speed-wise. I'd argue that this portion of cyclists may well do a disproportionately large number of miles, leave aside the way they do them.

Many other road risks are calculated per mile; it may turn out that these racey types are at greater risk just because of the mileage they do.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

pwward
Posts: 193
Joined: 12 Jan 2007, 10:48am
Location: Newcastle u Tyne

Re: New research on relative risks of different modes.

Postby pwward » 13 Dec 2012, 2:18pm

I'm very persuaded by CJ's argument

This recent research from Norways's Institute of Transport Economics would seem to add weight to the view that in places where cycling has become sportified, it's also becomes apparently more risky as utility cycling levels fall. I guess once helmets and day glow are added into the mix the cultural myth of cycling as dangerous is further re inforced.

http://www.sciencedirect.com/science/ar ... 7812000587

Ellieb
Posts: 899
Joined: 26 Jul 2008, 7:06pm

Re: New research on relative risks of different modes.

Postby Ellieb » 13 Dec 2012, 5:25pm

Alternatively the Dutch sports cyclists could be riding on dangerous things like roads, rather than on the cyclepaths where the utility cylists go. Who knows? :wink:

User avatar
CJ
Posts: 3119
Joined: 15 Jan 2007, 9:55pm

Re: New research on relative risks of different modes.

Postby CJ » 17 Dec 2012, 2:55pm

Brucey wrote:this may sound like a daft question, but how is (say) '1% of cycle traffic' calculated?

... it may turn out that these racey types are at greater risk just because of the mileage they do.

AFAIK it's from counting passing heads/helmets at a number of representative locations. That's the simplest, usual way of collecting traffic stats and automatically controls for any differences in exposure of one group versus another, since those who ride furthest are most likely to pass a counting point - or even more than one.

Edited to add:

If you've time to spare you can do your own survey by taking a random trip around the Netherlands by Google Streetview and counting the cyclists you see. This will be biased towards roads and cyclepaths near roads, rather than those that are out of sight of roads, so it'll probably be biased towards helmets, but I guarantee you won't see anything approaching 1 in 8 riders wearing them.
Chris Juden
One lady owner, never raced or jumped.